• Welcome to the Speedsolving.com, home of the web's largest puzzle community!
    You are currently viewing our forum as a guest which gives you limited access to join discussions and access our other features.

    Registration is fast, simple and absolutely free so please, join our community of 40,000+ people from around the world today!

    If you are already a member, simply login to hide this message and begin participating in the community!

Probability Thread

Renslay

Member
Joined
Aug 1, 2011
Messages
1,716
Location
Hungary
WCA
2005HANT01
YouTube
Visit Channel
*1/72 for the original pll skip = 1/373248

That is the probability for that particular three solves. But if you have a PLL skip, then the probability of having three in a row PLL skip (i.e., two more), is (1/72)^2. Conditional probability.

Edit:
Okay, technically, the "probability of having 3 PLL-skip in a row" has an important question: out of how many solves?
 
Last edited:

Renslay

Member
Joined
Aug 1, 2011
Messages
1,716
Location
Hungary
WCA
2005HANT01
YouTube
Visit Channel
Let's say I do N solves today. What is the probability of P = I have at least 3 PLL skips in a row?
Let p denote the probability of a single PLL skip.

N = 1 -> P = 0
N = 2 -> P = 0
N = 3 -> P = p^3
N = 4 -> P = (p^3)(1-p) + (1-p)(p^3) + p^4 (first 3 solves with PLL skip then no PLL skip -OR- first solve with no PLL skip then 3 PLL skip -OR- 4 PLL skips). Note that this is higher than p^3
...
N = k -> (okay, I'm too lazy for that)
...
N = inf -> P = 1

Or question number 2: what is the flaw in the reasoning above?
 

Ollie

Member
Joined
Mar 31, 2012
Messages
2,848
Location
London, UK
WCA
2012FROS01
YouTube
Visit Channel
Let's say I do N solves today. What is the probability of P = I have at least 3 PLL skips in a row?
Let p denote the probability of a single PLL skip.

N = 1 -> P = 0
N = 2 -> P = 0
N = 3 -> P = p^3
N = 4 -> P = (p^3)(1-p) + (1-p)(p^3) + p^4 (first 3 solves with PLL skip then no PLL skip -OR- first solve with no PLL skip then 3 PLL skip -OR- 4 PLL skips). Note that this is higher than p^3
...
N = k -> (okay, I'm too lazy for that)
...
N = inf -> P = 1

Or question number 2: what is the flaw in the reasoning above?

Not sure if you're deliberately trying to sound condescending, or what you're trying to prove, but the probability is the same regardless of how many solves you do in a day. If you flip a coin 15 times and it comes up heads on all of those, it does not make it more likely that you'll get a tails on the 16th time.
 

CubeRoots

Member
Joined
Mar 22, 2012
Messages
538
Location
Leicester, UK
WCA
2012LIVS01
YouTube
Visit Channel
That is the probability for that particular three solves. But if you have a PLL skip, then the probability of having three in a row PLL skip (i.e., two more), is (1/72)^2. Conditional probability.

Edit:
Okay, technically, the "probability of having 3 PLL-skip in a row" has an important question: out of how many solves?

he said, I just got 3 in a row. The probability that he would just get 3 in a row is 1/373248 as you know.

Your question is interesting but it doesn't change anything about the calculations that were made
 

mark49152

Premium Member
Joined
Oct 29, 2012
Messages
4,719
Location
UK
WCA
2015RIVE05
YouTube
Visit Channel
he said, I just got 3 in a row. The probability that he would just get 3 in a row is 1/373248 as you know.
Wrong. That is the probability for any three particular solves. The probability of having PLL skips on the first three solves of the day is 1/373248. The probability of having 3 in a row at some unspecified point in a session of many solves is obviously higher.
 

Ollie

Member
Joined
Mar 31, 2012
Messages
2,848
Location
London, UK
WCA
2012FROS01
YouTube
Visit Channel
Wrong. That is the probability for any three particular solves. The probability of having PLL skips on the first three solves of the day is 1/373248. The probability of having 3 in a row at some unspecified point in a session of many solves is obviously higher.

Irrelevant.

I just got three PLL skips in a row (Which also gave me my best single time, and average). Probability: 1/373248.

No mention of the average occurring at any specific point during a session, no mention that it was done in the first solves of the day and no mention of the number of solves done in the session. Therefore his calculation is correct. Stop over-complicating it.
 
Last edited:
Joined
Dec 18, 2012
Messages
61
Location
Earth
WCA
2012COOK01
No mention of the average occurring at any specific point during a session, no mention that it was done in the first solves of the day and no mention of the number of solves done in the session. Therefore his calculation is correct. Stop over-complicating it.

Thank you for clarifying. Three successive solves in a row are an independent event, and therefore the probability of three solves in a row being PLL skips is not affected by any prior or following solves.
 

Renslay

Member
Joined
Aug 1, 2011
Messages
1,716
Location
Hungary
WCA
2005HANT01
YouTube
Visit Channel
There is a difference between "the probability of having the same outcome three times in a row for the next three cases" (dices, coins, PLL skip, etc), and "there will be three identical outcomes in a row in the next K cases" The first one is clearly independent as you say, P = (1/p)^3, however, the latter has lim P = 1 as K -> inf, because it is not specifyed where will be that row in the series. Of course no matter how many cubes I already solved, a chance for a PLL skip (or 3 in a row) is still the same. But that is because we're talking about the next case. But if I'm talking about "what is the chance that I will have a PLL skip today after ten thousand solves", that is pretty high, because I did not specify which solve contains the skip.

That's why I asked how many solves was that day, because I was wondering the chance of having 3 PLL skips in a row anywhere during the solving.

Again, one who has cubing for years has much more chance that he had 3 PLL skips in a row than a fresh cuber. But speaking of the chance that someone will have 3 PLL skips in a row is the same.

Aaaand yes, I'm starting to over-complicate it. Sorry for that.
 

ryanj92

Member
Joined
Dec 26, 2011
Messages
1,575
Location
Sheffield, UK
WCA
2012JONE03
YouTube
Visit Channel
Surely the phrase 'if I do N solves what is the probability of getting 3 PLL skips in a row' has the following solution:

(N-2)*[(1/72)^3]*[(71/72)^N-3]
for N >= 3
The N-2 factor comes from the fact that the three PLL skips can be solves 1, 2, and 3, or 2, 3, and 4, right up to N-2, N-1, and N. This gives N-2 possibilites. Then the individual probability (in the knowledge that each solve is independent of the last solve) is the probability of one PLL skip cubed, multiplied by the probability that all other solves are not PLL skips.

EDIT: I've seen my own flaw already - my model assumes that ALL other solves have to not have a PLL skip, whereas you should include the other cases really (ie in 5 solves, solves 1-3 are skip, 4 is not skip and 5 is skip). in reality this diifference should be very small as probability of PLL skip < probability of non-PLL skip) so the above should give a good estimation?
 
Last edited:
Top